Jul 2026· Annual International Computer Software and Applications Conference· pp. 3098-3107· 0 citations· 54 references
Abstract
Large Language Models (LLMs) are now deployed at an unprecedented scale across many critical sectors, rapidly transitioning from experimental AI tools to embedded components of production software systems. This accelerated adoption, often enabled by low-code integrations, has lowered technical barriers while simultaneously expanding the attack surface of modern applications, particularly when deployments occur without sufficient domain-specific security expertise. In many cases, security maturity has not progressed at the same pace as capability expansion, creating systemic exposure across confidentiality, integrity, and availability dimensions. To provide structured clarity amid this rapid growth, this paper presents a comparative and standards-aligned analysis of LLM security risks and defense mechanisms grounded in the OWASP GenAI Top-10 (2025). We systematically examine each vulnerability class, map representative attack patterns to primary mitigation strategies, evaluate their security property impact, and analyze practical limitations and implementation trade-offs. In addition, we introduce a severity-based assessment to prioritize risks according to operational and systemic impact, offering a quantitative perspective on defensive readiness. Our findings indicate that current mitigation strategies are predominantly reactive, concentrated at inference time, and unevenly distributed across the LLM lifecycle. Controls addressing training pipelines, supplychain dependencies, and autonomous system behaviors remain comparatively less mature and less standardized. By integrating vulnerability classification, defense mapping, severity prioritization, and trade-off analysis within a unified framework, this study provides actionable guidance for strengthening secure, resilient, and standards-driven LLM deployment in high-stakes environments.
Automated vulnerability repair with large language models is increasingly woven into security workflows, yet evaluation practice has focused almost entirely on a single question: does the patch eliminate the targeted weakness? However, an equally important question has received little systematic attention: has the patching act itself introduced new weaknesses into the function being repaired? Bypass testing and regression measurement are orthogonal properties, and neglecting the latter leaves a material blind spot in any assessment of patch quality.This paper presents an empirical study of vulnerability regression in LLM-generated security patches, applying differential static analysis to 479 complete C/C++ functions spanning eight CWE types. Patches are generated from three frontier models alongside developer-authored ground-truth fixes, for a total of 1,916 analyzed patch–function pairs. AI models fix targeted vulnerabilities at rates approaching human developers, yet they leave a meaningfully larger share of patches in a structurally neutral state, modifying code without engaging its underlying weakness. Across all three AI models, regression rule sets are more similar to each other than to human regression patterns, a consistency specific enough to target with shared automated validators. Finally, for human developers, successfully fixing a vulnerability is strongly associated with lower regression risk. However, this association is absent and not statistically detectable for AI patchers, a distinction with direct implications for how AI patch quality should be verified.
Yasmeen Rawajfih, Adiba Mahmud, Ross Arnold· International Conference on...· 0 citations
Large language models are increasingly used to author Infrastructure-as-Code (IaC), where a single insecure default can be deployed directly into production. Prior evaluations report raw vulnerability counts for model-generated IaC, but without a human baseline they cannot determine whether models are actually worse than engineers. We introduce GenIaC-SecBench, a benchmark of 100 deployment scenarios stratified by architectural complexity, evaluated across 12 model configurations from four vendors, producing 1,196 IaC artifacts scanned by three independent policy engines (Checkov, Trivy, KICS). Critically, we also scan 634 human-authored IaC templates with the same toolchain, providing the first size-matched human security baseline. Vulnerability density is strongly inverse to artifact size (Spearman $\rho = -0.55$, $p<10^{-77}$), meaning unmatched comparisons measure size rather than security. When matched on declared-resource count, all model configurations fall within 3.21x--3.87x the human vulnerability density, with the gap widening for simpler tasks (4.9x at one resource, 1.4x at twenty or more). We decompose reasoning into standard generation, prompt-engineered chain-of-thought, and vendor extended-thinking APIs. Vendor extended thinking significantly outperforms prompted chain-of-thought ($-12.0\%$, $p = 0.0013$), while prompted chain-of-thought is indistinguishable from standard generation ($-1.3\%$, n.s.). Token instrumentation shows extended thinking uses under 1\% of the output budget, explaining its bounded effect. Two negative results also emerge: deployability does not correlate with vulnerability ($r = 0.158$, $p = 0.625$), and classical complete-case Friedman testing is infeasible for realistic benchmark designs, motivating the Skillings-Mack statistic. All code, data, and regeneration scripts are released.
This work systematically generates syntactic variants of security-relevant code generation prompts and evaluates their impact on code security across multiple open LLMs and programming languages, identifying prompt syntax as a concrete security control surface and providing actionable guidance for reducing vulnerability risk in LLM-assisted development.
Matteo Cicalese, Antonio Della Porta, Stefano Lambiase et al.· 0 citations
The risk is not tied to any particular model but rather to the task category, and organizations should ask not which tool to trust, but instead whether LLM-generated automation code should be deployed without review.
Software security has been a long-standing and prominent topic in both industry and academia. However, with the increasing deployment of smart devices across various architectures, there is now a significant demand for cross-architecture software. For instance, the Heartbleed vulnerability (CVE-2014-0160), classified as CWE-125 (Out-of-bounds Read) and disclosed in the OpenSSL library in 2014, serves as a prime example of a widely-deployed software vulnerability capable of affecting systems across diverse computing environments due to OpenSSL’s broad adoption. More broadly, memory-related vulnerabilities remain a persistent and significant threat, accounting for a substantial proportion of reported CVEs in recent years. Moreover, performing static cross-architecture vulnerability analysis on binary code poses particular challenges due to the variations in instruction sets across different architectures. In this paper, we present a novel static approach, called BinCros, to identify known vulnerabilities across different optimization levels and CPU architectures in binary code. It mainly consists of three phases: 1) Generating ground-truth intermediate representation (IR) data from binary code based on domain knowledge. To address the cross-architecture problem, we consider IR, which is a set of smaller, lower-level instructions that break down, describe, and capture all the effects of an instruction from standard architectures like x86, ARM, and MIPS; 2) Learning good embeddings based on the cross-architecture ground-truth datasets. We employ a code-code learning method to capture the semantic differences and maximize the distribution divergences between vulnerable and non-vulnerable samples; 3) Building a prediction model based on high-level feature representations. To demonstrate the effectiveness of BinCros, we conducted experiments with a series of baselines including single architecture-based techniques, cross-architecture-based techniques, and so on. Experimental results show that BinCros outperforms the baselines by at least 15% in terms of F1-measure in almost all cases. We believe this work will inspire other researchers to consider natural language processing-related techniques for cross-architecture software vulnerability detection.
Shigang Liu, Di Cao, Chao Chen et al.· IEEE Transactions on Informa...· 0 citations
- The rapid explosion of web application capabilities over the past ten years has fundamentally redesigned how applications are delivered, simultaneously introducing an intricate, multifaceted attack surface that continues to evolve. Standard vectors, long thought to be understood — such as SQL Injection (SQLi), Cross-Site Scripting (XSS), and Cross-Site Request Forgery (CSRF) — still exist as potent hazards, yet they are increasingly overshadowed. Emerging cloud-native architectures, serverless delivery mechanisms, microservices, and AI-driven automation introduce entirely new categories of subtle, deeply embedded vulnerabilities. This research evaluates how these threats have metastasized and traces the origins of modern security vectors to determine if established defensive protocols remain effective against increasingly complex modern exploitation tactics. We conducted a strict Systematic Literature Review (SLR) structured by PRISMA 2020 guidelines, filtering an extensive initial pool of 885 records down to 62 core sources published between 2015 and 2026. This foundational dataset synthesizes observations from 44 peer-reviewed empirical studies alongside analysis from 18 key cybersecurity frameworks and official threat intelligence reports, incorporating guidance from standards bodies including OWASP, NIST, and ISO/IEC. The synthesized evidence reveals a definitive and strategic maturation in adversarial approach: threat actors are abandoning isolated, single-vulnerability exploits. They are instead executing prolonged, multi-stage campaigns that specifically leverage the trust relationships found in interconnected software ecosystems. While SQLi, XSS, and authentication weaknesses remain critical and frequent (identified within our 12 primary attack categories), a steep rise in complex, multi-stage exploit chains, AI-assisted reconnaissance, API breaches, and software supply chain compromises represents the new operational normal for adversaries. Furthermore, our analysis indicates that traditional defensive frameworks like secure development lifecycles, Zero Trust Architecture, DevSecOps, and Web Application Firewalls (WAFs) are no longer sufficient in isolation. Their mitigation capacity works only when supported by continuous, real-time context-aware monitoring and truly dynamic risk management, establishing an empirical baseline for architecting resilient security posturing that can keep pace with accelerating innovation.
Irene I. Eda, Jose Marcelito D. Brigoli, Teodoro B. Comayas Jr. et al.· Iconic research and engineer...· 0 citations